A frequent cause of concern among investigators performing quantitative RT-PCR is false positives caused by genomic DNA contamination of RNA preparations. Because PCR is such a sensitive technique, a single copy of a gene can, theoretically, be detected. Here we test various methods for removing DNA contamination. Genomic DNA false positive signals are easily identified by performing a "no-RT" control during RT-PCR. We can therefore assess the effectiveness of DNA removal methods by agarose gel analysis of the -RT reactions.
Three Methods of Removing Contamination from RNA
The data presented here results from testing several common methods of DNA removal from RNA samples. Each DNA removal technique was analyzed for effectiveness by PCR amplification of the RNA with and without prior reverse transcription. The methods include:
DNase digestion
DNase is an endonuclease that cleaves DNA by breaking phosphodiester bonds. It must be inactivated or removed from the reaction prior to PCR, otherwise, it may digest newly amplified DNA. For this study, we tested 2 concentrations of DNase I (10 and 50 µl DNase I/ml sample) and four DNase I inactivation/removal methods:
- Chelation with 20 mM EDTA
- Heating at 70°C for 5 minutes
- Proteinase K digestion followed by phenol/chloroform extraction and NH4OAc/EtOH precipitation
- Protein removal using Ambion's RNAqueousÌ Kit
Acid phenol:chloroform extraction
Acid phenol:chloroform (5:1 phenol:CHCl3; pH 4.7) extraction partitions DNA into the organic phase. The RNA remains in the aqueous phase and can be subsequently recovered by precipitation.
Lithium chloride (LiCl) precipitation
LiCl precipitation is a selective precipitant of RNA. It inefficiently precipitates DNA which is discarded in the supernatant.
Assessing DNA Contamination of RNA Samples
To confirm the presence of contaminating DNA, two RNA samples were assessed on a 1% denaturing agarose gel by ethidium bromide staining (Figure 1a). One RNA sample showed visible DNA contamination (sample B), while the other did not (sample A). The same samples were then subjected to RT-PCR, along with a "no-RT" control, for the presence of ribosomal protein S15 message. Regardless of whether contaminating DNA was visible by gel assessment, both RNA samples showed amplifiable DNA in the no-RT control (Figure 1b). Once DNA contamination was verified, both RNA samples were subjected to the three DNA removal methods described above.
|
 |
| Figure 1. Assessment of DNA Contamination in RNA Preparations. Figure 1a. 1 µg of each of two RNA samples (labeled A and B) along with 3 µg of Ambion's MillenniumÌ Markers was assessed on a 1% denaturing agarose gel and stained with EtBr. Notice in Sample B, the high molecular weight DNA contamination. Figure 1b. 1 µg of each RNA from Figure 1a was used in a reverse transcription reaction performed with the RETROscriptÌ Kit. 1 µl of each RT reaction (RT-PCR) and 0.5 µg of each RNA ("no-RT" control) were subsequently used in a standard PCR reaction for the amplification of the ribosomal protein S15 message. One-tenth of each PCR reaction was assessed on a 1% agarose gel and stained with EtBr. |
Results
DNase digestion
Both RNA samples A & B were treated with 10 and 50 U DNase I (Ambion, Inc.) per ml RNA sample at 37°C for 30 minutes. The digested samples were split into four tubes and each was treated with a different DNase inactivation or removal method. The data reveal that DNase treatment followed by any of the inactivation methods was sufficient to remove contaminating DNA as tested by the no-RT control (Figure 2a) and did not inhibit subsequent RT-PCR (Figure 2b).
Acid phenol:chloroform extraction
Aliquots of each RNA sample were extracted with an equal volume of acid phenol:chloroform followed by precipitation with 0.5M NH4OAc and EtOH. The RNA pellets were resuspended in nuclease-free water and then tested in RT-PCR reactions. As with DNase digestion, acid phenol:chloroform extraction removed contaminating DNA without effecting RT-PCR.
LiCl precipitation
RNA samples A and B were precipitated with equal volumes of 7.5M LiCl, resuspended in nuclease-free water and used in RT-PCR reactions. LiCl precipitation did remove DNA from sample A but was insufficient to remove the greater DNA contamination in RNA sample B (Figures 2a and 2b).
|
 |
Figure 2. Assessment of Methods to remove DNA Contamination of RNA. The two RNA samples shown in Figure 1a were treated with DNase, extracted with acid phenol or precipitated with LiCl as a method of DNA removal. The RNase were treated with either 10 U or 50 U DNase I/ml RNA sample,split into four aliquots and the DNase I was inactivated by one of four methods:
Figure 2b. Approximately 1 µg of each treated RNA was used in a RT-PCR reaction for the amplification of ribosomal protein S15 message using Ambion's RETROscriptÌ Kit. One-tenth of each RT PCR reaction was assessed on a 1% agarose gel and stained with EtBr. |
Conclusions
Three common methods used to remove contaminating DNA from RNA preparations were tested. The methods included DNase I digestion, acid phenol:chloroform extraction and LiCl precipitation. Both DNase I digestion (with subsequent DNase I inactivation or removal) and acid phenol:chloroform extraction were sufficient to remove contaminating DNA as tested by no-RT control PCR reactions. These methods worked well even when the amount of DNA contamination was visible by EtBr staining of the RNA sample (sample B). Neither method inhibited subsequent RT-PCR reactions. LiCl precipitation as a method to remove contamination DNA from RNA samples was not as efficient. LiCl precipitation was adequate to remove moderate but not gross DNA contamination from RNA samples. LiCl precipitation also did not inhibit subsequent RT-PCR reactions.
